RAG-Chat

RAG-Chat is a powerful chatbot built using the LangChain library. It can decide whether to respond using preloaded document data or perform a real-time web search using the Brave Search API. This flexibility allows it to provide accurate and up-to-date information based on the user's queries.

Features

• Document Loading: Load documents from specified files, directories, or predefined paths.
• Decision Making: Uses the Ollama LLM to decide if a query can be answered using preloaded documents or if a web search is required.
• Web Search: Integrates with the Brave Search API to fetch real-time information when necessary.
• Context Management: Efficiently manages context for accurate and relevant responses.

Installation

1. Clone the repository:

   git clone https://github.com/Dizzguise/rag-chat.git
   cd rag-chat

2. Create and activate a virtual environment:

   python -m venv venv
   source venv/bin/activate  # On Windows use `venv\Scripts\activate`

3. Install dependencies:

   pip install -r requirements.txt

Configuration

1. API Keys: Ensure you have your Brave Search API key.

2. Configuration File: Edit the config.yaml file to include your API keys and other settings.

   logging:
     level: DEBUG
   
   llm:
     model: llama3
   
   web_search:
     api_key: YOUR_BRAVE_SEARCH_API_KEY
     count: 10
     extra_snippets: true

Usage

1. Run the chatbot:

   python main.py

2. Interact with the chatbot:

   • Upon starting, you will be prompted to choose a document loading method.
   • Type your queries and receive responses based on preloaded documents or real-time web search results.
   • Type exit or quit to end the session.

File Structure

• main.py: The entry point of the application. Handles user input and coordinates the decision-making and response generation.
• config.yaml: Configuration file for API keys and settings.
• document_loader.py: Handles loading documents from specified files or directories.
• retrieval_chain.py: Creates the retrieval chain for document-based responses.
• decision_agent.py: Decides whether to use document data or perform a web search.
• query_agent.py: Handles querying the document vector store.
• context_agent.py: Manages context for generating responses.
• web_search.py: Integrates with the Brave Search API to perform web searches.

License

This project is licensed under the MIT License. See the LICENSE file for details.

Contributing

1. Fork the repository.
2. Create a new branch (git checkout -b feature-branch).
3. Commit your changes (git commit -am 'Add new feature').
4. Push to the branch (git push origin feature-branch).
5. Create a new Pull Request.

Contact

For any questions or feedback, please open an issue on GitHub.

---

Thank you for using RAG-Chat! We hope it provides valuable assistance in your projects.